Proceedings of the 2014 Winter Simulation Conference A. Tolk, S. Y. Diallo, I. O. Ryzhov, L. Yilmaz, S. Buckley, and J. A. Miller, eds. INVESTIGATING THE HIDDEN LOSSES CAUSED BY OUT-OF-SHELF EVENTS: A MULTI-AGENT-BASED SIMULATION Priscilla Avegliano Carlos Cardonha IBM Research Rua Tut´ oia, 1157 S˜ ao Paulo, SP CEP 04007900, BRAZIL ABSTRACT Out-of-shelf events refer to periods of time in which items of a certain product are not available to customers. It is clear that incidents of this nature result in economic loss, but their side effects are much more profound: since there is no record of missed sales opportunities, the estimated demand curve tends to be inaccurate. As a result, order placement strategies employed by retailers are based on imprecise forecast models, so further out-of-shelf events are very likely to occur: a vicious cycle, hence, arises. In this work, we propose a multi-agent-based simulation to evaluate the impact of out-of-shelf events that considers the reactions of customers towards these incidents and retailers’ ordering strategies. Our results show that these events have a significant effect on demand estimation and that multi-agent-based simulations may provide interesting insights and support for the development of more accurate forecast models in retail. 1 INTRODUCTION Demand estimation is one of the most important challenges faced by the retail industry. Motivated by the unquestionable economical relevance of the problem, researchers and practitioners have been dedicating a considerable amount of effort to identify accurate forecast models. However, there is still a lack of satisfactory solutions for certain real-world scenarios (Papakiriakopoulos, Pramatari, and Doukidis 2009). Typically, statistical models are imprecise when they fail to consider relevant parameters of the problem, and in this context, the impact of out-of-shelf events in fluctuations of product demand has been one of these overlooked aspects. The term out-of-shelf (OOS) (or on-shelf unavailability) is used when a certain product is not available to customers on the shelf at a retail store. Despite the substantial losses that it inflicts on retailers and manufacturers, the level of on-shelf unavailability remains steady at around 8% worldwide (a value that grows up to 12% for items on sale) since the 90’s (Gruen, Corsten, and Bharadwaj 2002). OOS is difficult to prevent because it can be triggered by a multitude of root causes; obviously, when a retailer runs out of items in the back-room stock (which can be a consequence of poor demand forecasts or phantom stocks), the problem is unavoidable; moreover, on-shelf unavailability also takes place if replenishment plans are misaligned with sales velocity; finally, product misplacement is another serious related issue, caused both by planogram’ nonconformities (e.g., employees do not put the items on the correct shelf) and by customer interaction (e.g., someone takes an item and abandon it in an incorrect shelf). The impact of OOS events for retailers is probably more complex than a superficial analysis would allow one to see, since it has several direct and indirect consequences. It is estimated that on-shelf unavailability is the cause of 3.9% to 4.5% sales losses (Gruen, Corsten, and Bharadwaj 2002), so it is clear that the most immediate effect is economically significant. However, in addition to that, we should also take into account that customers facing OOS events may have different types of reactions; while some people postpone their 242 978-1-4799-7486-3/14/$31.00 ©2014 IEEE